Round 1

Reviewer 1 Report

The manuscript "Fine mapping of cropping intensity in complex planting areas using phenology algorithm, Sentinel-2, and Google Earth Engine" describes an approach based on automated mapping of crops into three cathegories based on the threshold approach. The authors used Google earth engine and Sentinel 2 spectral images to achieve their goal. LSWI and NDVI were chosen as phenological metrics.
In its current form, the manuscript resembles technical notes rather than a full-fledged article.
No methodological know-how is described, only a diagram of the sequence of operations is given. 
1. The reviewer strongly disagrees with the need to use filters to correct the original images - all the necessary steps have already been taken at the stage of selecting the level of image processing (SR), masking clouds and shadows. Because of the filters applied, the original data are subjected to changes, which raises doubts about the reproducibility of the experiment. 
2. Why were only selected vegetation indices used? 
3. How were the visible and infrared data corrected for SWIR? They have different spatial resolutions. Which SWIR was used? Since the SWIR channel has a resolution of 20 m, then what final resolution of the final product did the authors get?
4. The main questions are related to the statistical unprovenness of the performed works. The work based on the threshold approach uses multi-year data, taking into account the variance, which allows to take into account climatic features, spatial patterns of growth of each crop. It is not known how these factors are taken into account by the author.
5. In addition to comparing the data on the calculated areas and the state statistics data, it is necessary to give the data on calculations by alternative, already proven methods. Since the authors used GEE, it is possible to make a classification using Random forest method or use other methods. In addition, to analyze the results of the comparison of areas, you must provide tables, analysis, histograms, etc., to which the main errors and whether they have a systematic component.
In view of all of the above, I consider that the authors should finalize the study, taking into account the comments.

Author Response

Dear Reviewers，

  Thanks very much for taking your time to review our manuscript entitled “Fine mapping of cropping intensity in complex planting areas using phenology algorithm, Sentinel-2, and Google Earth Engine”(ID: ijgi-1303267). Those comments are all valuable and very helpful for revising and improving our paper, as well as the important guiding significance to our researches. We have studied comments carefully and have made correction which we hope meet with approval. Revised portion are marked in red in the paper.The specific response is in the file below.

  Yours sincerely,

Author Response File: Author Response.pdf

Reviewer 2 Report

The paper is generally well written and well structured. The phenology algorithm is clearly described, and the paper is based on tried-and-tested (but not cutting edge) indices, compositing and smoothing approaches. There are a few inconsistencies to check:

Depersonalise the introduction. L 43. In place of “Under the background of declining cultivated land area in our country”, try “Under the background of declining cultivated land area in the Peoples’ Republic of China (PRC)”. L 68. Consider a linking sentence. Crop intensity mapping over any sizeable region is a challenge. In this paper we consider regional scale crop intensity mapping by studying Henan Province in PRC (details on this region are given in the methods section below).

L57. Landsat revisit is 16 d for one satellite… there are very often two in operation in offset orbits bringing revisit to 8 d, and sidelap between orbits can bring revisit down even further. The S2 revisit you report (5d – 10d) is because of two satellites. You should therefore put 8d – 16d for Landsat.

L70. You state “Firstly, due to the influence of the atmosphere, clouds, cloud shadows, etc., data are often continuously missing, and the number of high-quality images varies greatly in time and space.” The data are not missing in these instances, it is simply you cannot see the area of interest on the ground because of the clouds and atmospheric effects. A meteorologist would not find the data to be missing.

L97. Define GEE the first time you use it in the main text.

L145 – 146. How accurate is the FROM GLC10 mask in this region?

L160 – 161. You will have to provide some detail on how you filled the gaps… from later text, I guess it was Maximum Value Composite?

L189 – L190. Smoothing incorporates noise into time series. You should consider using more advanced methods, such as Viovy’s Best Index Slope Extraction method. You should at least mention more advanced methods in your discussion and conclusions, even if you don’t re run the analysis using these improved approaches.

L243 – 245. How were these sample points allocated? What was the sampling scheme? How are they distributed spatially? How big were the sample points? (again, it is about providing all the information someone would need to replicate your experiment.

Author Response

Dear Reviewers，

  Thanks very much for taking your time to review our manuscript entitled “Fine mapping of cropping intensity in complex planting areas using phenology algorithm, Sentinel-2, and Google Earth Engine”(ID: ijgi-1303267). Those comments are all valuable and very helpful for revising and improving our paper, as well as the important guiding significance to our researches. We have studied comments carefully and have made correction which we hope meet with approval. Revised portion are marked in red in the paper.The specific response is in the file below.

  Yours sincerely,

Author Response File: Author Response.pdf

Reviewer 3 Report

Reviewer

ISPRS – International Journal of Geo-Information

Manuscript Number: ID 1303267

Title: Fine mapping of cropping intensity in complex planting areas using phenology algorithm, Sentinel-2, and Google Earth Engine

As requested, I have reviewed the above-titled paper for potential publication in the ISPRS International Journal of Geo-Information- MDPI Journal. I divided my comments in the sections presented as follows.

Before presenting such comments on the manuscript itself, I would suggest the authors to reflect about a change in the proposed title, since the proposal is related to the development of an algorithm applied to fine mapping of cropping intensity in complex planting areas.

Therefore, such proposal should be emphasized and should not be at the same level of importance when referring to Sentinel-2 and to Google Earth Engine as is stated in the title (“using phenology algorithm, Sentinel-2, and Google Earth Engine”). It could also be worthwhile to mention the case study conducted for Henan Province in China.

 For example, one could think about something that goes  like: ”Development of a new phenology algorithm for fine mapping of cropping intensity in complex planting areas using Sentinel-2 and Google Earth Engine – Case study: Henan Province in China”.

Contribution

This paper proposes the development of new phenology algorithm for fine mapping of cropping intensity. More specifically, the algorithm is proposed to identify areas with different crops and corresponding agricultural production and also different peaks related to biomass production along the year. In this sense, the area could have one peak, two peaks and even three peaks associated with a defined type of crop.

Thus, it is posed a challenge to identify such agricultural intensity production as we examine available sensors. The authors present some criticisms towards Landsat and MODIS datasets. The authors defend the use of optical images retrieved from Sentinel -2 A/B imagery due to its relatively high repetitive overpass orbit over a certain area  (5-10 days) and 10-m spatial resolution in contrast to Landsat and MODIS datasets. They made an option to examine all available images along the 2019-2021 time span in the Google Earth Engine platform. The proposed time span included winter crops from 2019 to 2020 and 2020 to 2021.  On the other side, the authors have not used Sentinel-1 dataset, which provides radar images, which are not so affected by clouds as optical images. Such radar images were addressed by the authors just as final comment for further investigation in the future. They made an option to use just optical images and procedures based on defining maximum value composition, filtering and implementing spatiotemporal linear interpolation in the case of missing data.

Moreover, the authors chose to study the Henan Province in China, which is recognized to produce about 10% of the total grain production in the country. In addition, the authors used basically two indexes, NDVI (normalized difference vegetation index) and LSWI (land surface water index) for discrimination of cropping mapping and cropping intensity.

They claim that the research results are able to support the evidence of the good performance of the proposed methodology. In addition, the authors foresee the extended use of the adopted approach.

I found the manuscript has an interesting goal to be pursued, however the text should be revised. The text needs to address more broadly the different types of algorithms for crop classification and crop intensity, including indexes, jointly with a broad and updated literature review in this respect to make more clear the original contribution of the manuscript. Assumptions should be clearly presented along the manuscript. There are some gaps along the text with respect to the methodological approach and evaluation of the datasets used. I  do recommend  that the authors revise the text, including a bit of rewording and revision of the English language, even though it is generally fine to read.  There is also a need to produce more results and including more Figures and Tables to better explain and justify the methodology used. Some further comments are going to be raised in the next section with respect to the points raised.

Therefore, the more specific comments and questions regarding the manuscript paper are going to be presented jointly with the evaluation of the contribution of the manuscript in the next section. They are provided below with more details with respect to the methodological approach. I think it would be interesting to have some feedback from the authors and also to have fully revised version of the text in order to better refer to the points I will raise in the next paragraphs. That might also lead to explore or reflect about different scenarios still not well and thoroughly explored by the authors in the proposed manuscript but that deserves attention.

Please, see further comments in the attached file. 

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The authors have seriously revised the manuscript according to the comments.
After a final reading and editorial revisions, the manuscript may be published. 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Reviewer

ISPRS – International Journal of Geo-Information

Manuscript Number: ID 1303267

Title: Fine mapping of cropping intensity in complex planting areas using phenology algorithm, Sentinel-2, and Google Earth Engine

New Title: Development of a new phenology algorithm for fine mapping of cropping intensity in complex planting areas using Sentinel-2 and Google Earth Engine

As requested, I have reviewed the revised version of the above-titled paper for potential publication in the ISPRS International Journal of Geo-Information- MDPI Journal. There was a change in the title as suggested. I agree with the authors the new title is a good choice and better reflects the research work conducted.  I divided my comments in the sections presented as follows.

Contribution

This paper proposes the development of new phenology algorithm for fine mapping of cropping intensity. More specifically, the algorithm is proposed to identify areas with different crops and corresponding agricultural production and also different peaks related to biomass production along the year. In this sense, the area could have one peak, two peaks and even three peaks associated with a defined type of crop.

Thus, it is posed a challenge to identify such agricultural intensity production as we examine available sensors. The authors present some criticisms towards Landsat and MODIS datasets. The authors defend the use of optical images retrieved from Sentinel -2 A/B imagery due to its relatively high repetitive overpass orbit over a certain area  (5-10 days) and 10-m spatial resolution in contrast to Landsat and MODIS datasets.

They made an option to examine all available images along the 2019-2021 time span in the Google Earth Engine platform. The proposed time span included winter crops from 2019 to 2020 and 2020 to 2021.  On the other side, the authors have not used Sentinel-1 dataset, which provides radar images, which are not so affected by clouds as optical images. Such radar images were addressed by the authors just as final comment for further investigation in the future. They made an option to use just optical images and procedures based on defining maximum value composition, filtering and implementing spatiotemporal linear interpolation in the case of missing data.

Moreover, the authors chose to study the Henan Province in China, which is recognized to produce about 10% of the total grain production in the country. In addition, the authors used basically two indexes, NDVI (normalized difference vegetation index) and LSWI (land surface water index) for discrimination of cropping mapping and cropping intensity.

They claim that the research results are able to support the evidence of the good performance of the proposed methodology. In addition, the authors foresee the extended use of the adopted approach.

I found the manuscript posed an interesting goal to be pursued. In this second version of the manuscript, the authors presented adequately and clearly the scope of the research and its corresponding contribution jointly with a more complete set of references. In addition, proper exposition of the methodological framework was followed. Figures and results in general are more nicely disposed to the reader. 

Please, see further comments in the attached file. 

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

The authors used Sentinel-2 images available on the Google Earth Engine platform to develop a phenological-based algorithm to map cropping intensity in Henan Province (6,444 km²) for 2020. They used remote sensing data (Sentinel-2A and Sentinel-2B) during the 2020 annual period to compute the NDVI and LSWI time series and assess cropping intensity. The validation was processed using mapping accuracy (PA), user accuracy (UA), overall accuracy (OA), and Kappa coefficient.

Overall, the paper covers an important and necessary topic in analyzing and mapping cropping intensity. However, I feel this paper needs to provide a more critical discussion related to the cross-disciplinary scope of the journal Land.

The methods and results are adequately described. However, there is no discussion concerning the context of land system science. In my opinion, this piece would be better fitted in the Remote Sensing journal as a technical note.

Kind of questions that may be addressed:

- Why is the methodology applied just for 2020? If authors are using the GEE, why not provide a better understanding of the cropping intensity patterns in China over time, at least from 2015-2020?

- What are the agents, structures, and nature of cropping intensity in China and Henan Province? What is the implication for food security?

- How or to what extent the ecosystem services are affected by the cropping intensity?

- How population/economy/geopolitics may affect the cropping intensity dynamics?

- How do cropping intensity affect biogeochemistry, biodiversity, biophysical properties, and disturbance regimes of the terrestrial ecosystem?

- How the cropping intensity is linked to the local population?

So, my recommendation is to reject this manuscript, as I feel this piece does not fit the Land scope.